Tải bản đầy đủ - 0trang
6 A Coherentist Account of the Relation Between Acceptance and Acceptability: Reflective Equilibrium
A Coherentist View on the Relation Between Social Acceptance and Moral…
foundational and unchangeable moral values, they stress that moral judgments are
What is more, coherentist approaches not only allow for the possibility of revision of moral judgments, they may also point out how such revision might occur. In
a coherentist account moral revision is not based on a sudden struck of insight but
rather it implies a revision of the entire web of someone’s moral judgments, or at
least a relevant part of it. Revision may for example begin with certain tensions
between moral beliefs which may trigger changes in the entire web, so achieving
new coherence (and new justification). This picture seems to strike a healthy balance between openness to moral revision and a certain conservatism when it comes
to well-established moral norms and values (because revision in principle relates to
the entire web of moral judgments, which will not be so easily overturned).
Rawls’ model of (wide) reflective equilibrium may be seen as a specification of
what a coherentist account of ethics could look like and how it allows for moral
revision. Rawls’ model distinguishes three layers of moral judgement: considered
moral judgements, moral principles and background theories. These layers do not
represent distinctions in moral force (or justification), but rather different levels of
abstraction or generality. Background theories are usually the most general and
abstract, moral principles are somewhat more specific and considered moral judgments are judgements about specific cases or situations.
Although revision may start at all three layers, it would seem that in the case of
established moral belief systems, it will often start with new considered judgments.
This is the case because considered judgments relate to specific situations or cases,
and new situations may lead to new considered judgments. One important reason
for this is that considered judgments are not only action-guiding, i.e. they do not just
derive from moral theories and principles, but are also world-guided, i.e. they are
also based on the facts of the specific situation or case.
If we apply this idea to our discussion about the relation between acceptance and
acceptability of technology, we get something like the following picture. People’s
considered judgments about the acceptability of a specific technology (in a specific
situation) may on the one hand been seen as ‘facts’ about the situation (i.e. the fact
that people judge so and so). As we have seen, these facts are relevant for whether a
technology is (actually) accepted. At the same time, such factual judgments feed
into the judgement whether a technology is acceptable. The acceptability judgment,
however, requires coherence between all three layers of the wide reflective equilibrium model. If someone’s initial (considered) judgement is not coherent with his or
her other moral beliefs, such coherence may be achieved either by revising the judgment or by revising other elements in his or her entire web of moral beliefs.
One way to understand the distinction between acceptance and acceptability in
the light of Rawls’ model of reflective equilibrium is to make use of the distinction
between narrow and wide reflective equilibrium (Daniels 1979; Rawls 2001). In a
narrow reflective equilibrium, we typically look for the moral principles and theories that fit our actual, given moral judgment best and we do not critically scrutinize
our judgments about particular cases in the light of a range of moral theories and
principles. Narrow reflective equilibrium is then more a descriptive mode and it
I. van de Poel
reveals the moral ‘grammar’ that underlies actual moral judgments. De facto judgments that establish the acceptance of a technology seem to be based on such a
narrow reflective equilibrium. In a wide reflective equilibrium, we more broadly
reflect on our moral judgments, viewing them from a number of moral perspectives
and principles, and striving for full coherence rather than just adapting our theories
and principles to given judgments. Contrary to a narrow reflective equilibrium that
is mainly descriptive, a wide reflective equilibrium justifies moral judgments. So,
judgments about acceptability appear to require a wide reflective equilibrium
whereas acceptance requires only a narrow equilibrium.
Especially in his later work, Rawls allows for the possibility that different people
might come to different wide reflective equilibria (Rawls 2001). He considers this
possible because people may have different world views, and even if the wide reflective equilibrium approach requires one to consider a range of moral perspectives
and principles, it does not require accepting all these perspectives and principles and
including them in the agent’s web of moral beliefs. In fact, it would seem impossible to include all possible moral perspectives and principles into one coherent web
of moral beliefs. In this sense, even if a wide reflective equilibrium has a certain
justifying power, it is not necessary that people come to the same wide reflective
equilibrium and therefore to the same judgment for example on the acceptability of
Still, Rawls believes that for many public issues a so-called overlapping consensus is possible. An overlapping consensus is a moral belief (principle, judgment
about a case, etc.) that is shared among the reflective equilibria of different people.
Rawls believes that such an overlapping consensus is possible if not likely for those
public issues for which people are willing to reason from public reason; i.e. on the
basis of moral reasons that are relevant for the public domain and that accept what
he calls reasonable pluralism, which include beliefs like that the state should not
force people to accept certain nonpublic values and that it is possible to draw a
boundary between public and nonpublic values.
The wide-reflective-equilibrium model and the notion of overlapping consensus
can thus account for the (potential) difference between acceptance and acceptability, as well as for their relation. Acceptance can be understood as referring to people’s de facto judgements in a narrow reflective equilibrium, while acceptability
refers to the judgment that people would have if they would achieve a wide reflective equilibrium. Moreover, as we have seen, people’s judgments on acceptability
that are the result of achieving a wide reflective equilibrium by that person may still
vary between persons. However, one might expect, or at least hope, that if people
are willing to reason from public reason, an overlapping consensus will develop that
represents a justified consensus on the acceptability of a technology.
One might then argue that according to a coherentist account, acceptance and
acceptability coincide (or should coincide) in the ideal case. The reason for this is
that in the case of full coherence (or wide reflective equilibrium), people’s (de facto)
judgments about specific cases are in full coherence with the entire web of their
moral beliefs, and are based on a wide reflective equilibrium. So their de facto judgments (acceptance) are justified due to the coherence with all their other moral
A Coherentist View on the Relation Between Social Acceptance and Moral…
beliefs and hence also express acceptability. In addition, in the ideal case that people
reason from public reason, their judgements on acceptability will overlap, so representing an overlapping consensus on the moral acceptability of a technology.
In practice, however, this ideal situation is not very likely to be (completely) realized. First, there may and often will be differences between people’s de facto judgments and the judgments they would have in wide reflective equilibrium. One reason
for this is that full coherence is more a regulative ideal than a practical possibility.
It does not seem unrealistic to assume that people’s moral belief systems, just as
their cognitive belief systems, always have some degree of lack of coherence.
Although coherentist accounts of ethics postulate coherence as an ideal, it should be
noted that an actual lack of coherence is not always bad (also not in the light of the
ideal), as it can be a source of moral reflection, scrutiny and revision. In other words,
lack of coherence may lead to moral reflection and that often seems a good thing.
Second, even if people’s individual judgments are based on a wide reflective
equilibrium, that is not yet a guarantee for an overlapping consensus with respect to
the acceptability of a technology. Again, however, this lack of consensus need not
necessarily be seen as only problematic, because it may also be a source of debate,
argumentation and reflection. In other words, it would seem that reasonable people do
not only form and adapt their wide reflective equilibrium as the result of intrapersonal
deliberation but also as the result of deliberation with other persons. And again reasoning from public reason and the attainment of an overlapping consensus might be more
a regulative ideal rather than a possibility that can be easily obtained in practice.
As stressed before, it is therefore important not to equate too easily acceptance
and acceptability (even if they may coincide in the ideal case), because the distinction between the two invites moral reflection and debate.
In order to clarify the notions of acceptance and acceptability of technology and
their relation I first presented some preliminary explications of both notions, after
which I discussed the logical gap between the two notions. The idea of a logical gap
has a certain attractiveness because it illuminates the differences between the two
notions and it points to the danger of equating acceptance with acceptability. Still, I
argued we also need to account for how both notions may be related. To this end, I
first argued that both acceptance and acceptability are thick concepts that contain
descriptive as well as normative elements. However, that does not mean that they
cannot be distinguished. Next, by applying the wide-reflective-equilibrium model
of John Rawls to the discussion, we have seen that we can conceive of acceptance
as the result of a narrow reflective equilibrium, i.e. an equilibrium in which moral
principles and background theories are adjusted to given considered judgments.
Acceptability on the other hand would be the result of a wide reflective equilibrium
that also critically scrutinizes considered judgements from a variety of moral theories and background principles. In contrast to a narrow reflective equilibrium, a
I. van de Poel
wide reflective equilibrium also provides a moral justification and therefore has a
strong moral force. It might be argued that in the ideal case of a wide reflective
equilibrium with complete coherence, acceptance and acceptability coincide. This
is, however, more a regulative ideal than a practical possibility. Moreover, even in
wide reflective equilibrium, people might come to different moral judgments. They
may still be able to reach an overlapping consensus if they reason from public reason as suggested by Rawls, but again this is merely a regulative ideal.
For debates about the acceptability of a technology the coherentist view implies
that in judging the acceptability of technology we should take into account (rather
than ignore) the stakeholders’ acceptance (or non-acceptance) of a technology and
that non-acceptance should not be simply treated as a barrier to overcome. However,
the acceptance (or non-acceptance) by stakeholders need to be morally scrutinized,
reflected upon and brought into coherence with other moral beliefs in a wide reflective equilibrium, before it can also be seen as an expression of acceptability.
Moreover, people’s moral judgments should be a source of debate rather than be
seen as given and unchangeable.
More generally, a wide reflective equilibrium account, or for that matter other
coherentist accounts, does not assume a neat distinction between normative and
descriptive statements or judgments. In fact, at all three layers of the wide-reflectiveequilibrium model, we find both normative and descriptive elements. Among the
background theories we find not just moral theories but also descriptive theories,
and often they will be a combination of both; like for example utilitarianism which
is a moral theory but also a theory about human nature and about what makes people
actually happy. Moral principles, the second layer in the model, will often be formulated with the help of thick normative concepts; that is to say that they also have
some descriptive content. And as we have already seen before, considered moral
judgments are both action-guiding and world-guided.
Although we may still be able to distinguish analytically between the descriptive
and the normative, it may be difficult or impossible to unambiguously classify the
elements that are in coherence in a wide reflective equilibrium model as either
descriptive or normative. This also makes clear why we should take into account the
empirical or descriptive in applied ethics, as also advocated in the empirical turn in
the philosophy of technology, without assuming that normative or ethical questions
now get empirical answers.
Acknowledgments This paper is the result of various conversations with Peter Kroes about this
topic. I thank him wholeheartedly for these (and other) fruitful, inspiring and thought-provoking
discussions. Earlier versions of this paper were presented at the symposium on “Social acceptance
in energy through democratic and administrative procedures” held on November 6, 2014 in St.
Gallen (Switzerland) and at the SPT (Society for Philosophy of Technology) conference, held from
July 2–6, 2015 at Northeastern University in Shenyang (China). I also like to thank Maarten
Franssen for comments on an earlier version of the paper. Any mistakes in the paper are fully my
A Coherentist View on the Relation Between Social Acceptance and Moral…
Beauchamp, T. L., & Childress, J. F. (2013). Principles of biomedical ethics. New York: Oxford
Dancy, J. (2002). Practical reality. Oxford: Oxford University Press.
Daniels, N. (1979). Wide reflective equilibrium and theory acceptance in ethics. Journal of
Philosophy, 76(5), 256–282.
Emanuel, E. J., Wendler, D., & Grady, C. (2000). What makes clinical research ethical? The
Journal of the American Medical Association, 283(20), 2701–2711.
Hume, D. (2000). . A treatise of human nature. Oxford: Oxford University Press.
Moore, G. E. (1903). Principia ethica. Cambridge: Cambridge University Press.
Rawls, J. (1999). . A theory of justice. Cambridge, MA: The Belknap Press of Harvard
Rawls, J. (2001). Justice as fairness. A restatement. Cambridge, MA: The Belknap Press of
Harvard University Press.
Williams, B. (1985). Ethics and the limits of philosophy. Cambridge, MA: Harvard University
Zaal, M. P., Terwel, B. W., ter Mors, E., & Daamen, D. D. L. (2014). Monetary compensation can
increase public support for the siting of hazardous facilities. Journal of Environmental
Psychology, 37, 21–30.
Ibo van de Poel is Anthoni van Leeuwenhoek Professor in Ethics and Technology at Delft
University of Technology in the Netherlands. He has published on engineering ethics, the moral
acceptability of technological risks, values and engineering design, moral responsibility in research
networks, and ethics of new emerging technologies like nanotechnology. His current research
focuses on new technologies as social experiments and conditions for morally responsible
Perovskite Philosophy: A Branch-Formation
Model of Application-Oriented Science
Abstract In this paper, I present a model of application-oriented science, to supplement existing work in science and technology studies on the re-orientation of
scientiﬁc research. On this “branch-formation” model, research efforts may be
guided by non-epistemic values without compromising their epistemic value: they
may involve completion of mechanism representations that serve control over these
mechanisms while also adding to our understanding of them. I illustrate this model
with a case study from photovoltaic technology, involving the possible use of materials with the so-called ‘perovskite’ structure in dye-sensitized solar cells. The paper
has three parts. The ﬁrst argues how existing work on the increasing applicationorientedness of scientiﬁc research can and must be supplemented with a perspective
from the philosophy of science. The second presents the branch-formation model,
which combines central ideas of the ‘ﬁnalization-of-science’ program of the
Starnberg school with recent work in ‘mechanistic’ philosophy of science and in the
philosophy of technology. The third part illustrates the branch-formation model
with current developments in research on perovskite solar cells.
Keywords Basic and applied research • Mechanism • Mode 2 knowledge • Triple
Scientiﬁc research does not take place in a vacuum. Even those who strongly believe
in academic freedom and value-free science acknowledge that, in practice, research
is partly driven by societal needs, as well as by more mundane external forces such
as funding streams. Recently, however, many have claimed or voiced concerns that
scientiﬁc research, especially at universities,1 is increasingly focused on producing
knowledge or other results that are directly relevant to other stakeholders: industry,
governmental organisations, or society in general.
In this paper, I mostly use ‘research’ or ‘scientiﬁc research’, and leave implicit that this research
is traditionally (thought to be) done at universities.
W. Houkes (*)
Eindhoven University of Technology, Eindhoven, The Netherlands
© Springer International Publishing Switzerland 2016
M. Franssen et al. (eds.), Philosophy of Technology after the Empirical Turn,
Philosophy of Engineering and Technology 23,
There are several ways to characterise this change, depending on which aspect
one wants to emphasise, and how it is assessed. Some note a shift in the scientiﬁc
self-conception, towards a more “engineering” or “entrepreneurial” image (Daston
and Galison 2007, p. 382ff; Etzkowitz 2002). Others regard it as an alarming move
away from “pure” science, mainly under the inﬂuence of industry (Ziman 2000), as
a transition, or as a revolution. In this paper, I label the change as a re-orientation
towards applications. This label leaves it open whether researchers play an active
role in redirecting their efforts, or mainly react to external incentives. It also highlights, among all interrelated changes, that many research efforts in especially the
natural and life sciences are increasingly focused on devices and technologies. This
focus means that the re-orientation of research is of interest to both philosophers of
science and philosophers of technology, since it immediately concerns the interface
or ‘trading zone’ between their areas of interest.
The actual interests of especially philosophers of science hardly reﬂect the
ongoing changes in research practices. Researchers in science and technology
studies and in innovation studies have, by contrast, been very active in studying how
research is interwoven with industrial and societal concerns. In terms of publications in leading journals in the respective ﬁelds, there are more papers on the topic
in some single issues of Research Policy than in entire volumes of Philosophy of
Science.2 Likewise, in interdisciplinary projects devoted to recent changes in scientiﬁc research, philosophers are seldom involved.
That philosophers of science and technology have not paid much attention to the
re-orientation of scientiﬁc research does not, of course, mean that they should; nor
that, as latecomers to the discussion, they will have much to add. I will argue in this
paper that philosophers of science and technology can and need to contribute to
studying, and in particular assessing, the application re-orientation of scientiﬁc
research. Not to do so is an oversight, given the types of analysis that are currently
available. Furthermore, I show that philosophy of science and philosophy of technology offer materials to construct a supplement to existing approaches, and I illustrate the supplement with an example of research on photovoltaic cells, in particular
those containing materials with so-called ‘perovskite’ structure. This essay contributes to an empirical turn in the philosophy of science and technology: it uses and
develops approaches in these sub-disciplines to improve our insight into an ongoing
phenomenon – the application re-orientation of research – that has so far mostly
been addressed in science and technology studies. Furthermore, it is partly empirical in method: after an initial sketch, the branch-formation model presented in the
essay is developed in discussing a case study of ongoing application-oriented
Like perovskites, the paper consists of three parts. In Sect. 12.1, I review two of
the most inﬂuential existing approaches to the re-orientation of scientiﬁc research:
the ‘New Production of Knowledge’ programme and the ‘Triple Helix’ approach.
Some publications in philosophy of science and technology are concerned with the application
re-orientation, especially the growing role of commercial interests. Examples are Wilholt (2006),
Radder (2010), and Irzik (2010).
Perovskite Philosophy: A Branch-Formation Model of Application-Oriented Science
I argue that these leave room and, more strongly, create a need for a supplement that
looks at the contents of scientiﬁc research rather than its institutionalisation, and
that allows an evaluation of the epistemic merits of application-oriented research.
Philosophy of science and technology might offer the means to develop such a
supplement. In Sect. 12.2, I outline one possible supplement, which I call the
‘branch-formation’ model of application-oriented science. It is built on the central
tenets of the ‘ﬁnalization-of-science’ program, one of the few sustained efforts
within the philosophy of science to reﬂect on the epistemic legitimacy of orienting
scientiﬁc research towards broader, non-epistemic or ‘societal’ values. I show that
the central ideas of this programme can be modiﬁed with elements of mechanistic
philosophy of science and with work in the philosophy of technology on the design
of artefacts. This leads to a model that represents research efforts as branches that
develop representations of mechanisms, and that may share rudimentary versions of
these representations – mechanism sketches or functional analyses – with other
branches. Developments of branches and within branches may be guided by traditional scientiﬁc goals (i.e., creation of epistemic value3) through explaining and
predicting phenomena; by external, societal goals such as increasing safety and
sustainability through controlling phenomena in artiﬁcial devices; or by both types
of values and goals. As an illustration of the latter, I reconstruct, in Sect. 12.3, the
development of perovskite solar cells as a recent branch in photovoltaic research. In
this area, the same mechanism sketch (a three-part functional analysis) is shared by
several branches. Each branch seeks to complete the representation of the operating
mechanism of a particular type of photovoltaic cell, in order to enhance the performance and durability of this cell, while minimizing the cost of production; thus
combining scientiﬁc goals with societally or industrially relevant ones. Section 12.4
Existing Approaches and the Need for a Supplement
In this section, I argue that philosophers of science and technology can and, to some
extent, should contribute to studying the application re-orientation of scientiﬁc
research. I do so by reviewing existing approaches and showing that these leave
room and, more strongly, create a need for a supplement that may be developed by
philosophers of science and technology.
Out of the many approaches to understand the ongoing re-orientation of scientiﬁc
research,4 I focus on the “New Production of Knowledge” (NPK; Gibbons et al. 1994)
Throughout the paper, I sometimes use the distinction between (the creation of) epistemic and
non-epistemic values, and sometimes that between (the pursuit of) internal/scientiﬁc and external/
non-scientiﬁc (or ‘societal’) goals.
Often called ‘the science system’ or ‘systems’ in the science-studies literature.
and the “Triple Helix” (3H; Etzkowitz and Leydesdorff 2000).5 These are highly
inﬂuential6 and – despite their differences – point towards the same supplement.
After reviewing both approaches, I give two reasons why such a supplement is possible or even necessary. First, the abovementioned frameworks are largely descriptive, or prescriptive only insofar as they propose policies to facilitate the changes
that they describe. By contrast, philosophers could assess whether the phenomenon
constitutes a legitimate change of scientiﬁc research, or one that interferes with its
epistemic merits. Second, science-studies frameworks operate on a high level of
aggregation: they focus on changes in science as a whole, as a knowledge-producing
activity, and its place in society. By contrast, philosophers could focus, not on who
produces knowledge in which organizational form and under which social conditions, but on what knowledge is produced for which more proximate purpose.
Both approaches resulted from research programs and thematic conferences, and
harbour a variety of convergent views.7 The outlines given here ignore most of this
diversity for the sake of brevity.
The NPK approach revolves around a distinction between two ‘modes’ of knowledge production: the traditional ‘Mode 1’ and the new, emerging ‘Mode 2’. Mode 1
production primarily occurs within academic disciplines. These structure what
counts as ‘good science’, both in determining what is a valuable contribution (i.e.,
criteria of relevance, originality, etc.), and in providing systems of quality control
(in particular: review by scientiﬁc peers). In terms of physical location, Mode 1
production is ‘homogeneous’, because it is mostly a concern of research institutes
such as traditional universities; in terms of interrelations with other societal groups
or interests, it is ‘autonomous’, because it is not deeply or intrinsically concerned
with the impact of the produced knowledge outside academic disciplines. By contrast, Mode 2 knowledge production is a heterogeneous, transdisciplinary affair.
A larger variety of institutions – including government agencies, high-tech spinoff
companies and consultancy ﬁrms – is involved in the production of Mode 2 knowledge, which irreducibly combines disciplinary perspectives and concepts.
Furthermore, it is application-driven: knowledge production cannot be disengaged
from its practical applications, because these structure what is ‘good science’ just as
much as any disciplinary standard. Sensitivity to, for instance, the commercial or
societal value of knowledge is an integral part of Mode 2 knowledge production,
and such values enter into the quality control of its products – as universities in
many countries have noted in the changing terms of assessment of their research.
Hessels and Van Lente (2008) provide a more complete review of approaches, and focus on the
New Production of Knowledge.
The presentations of the NPK- and 3H approaches in this paper are based on Gibbons et al. (1994)
and Etzkowitz and Leydesdorff (2000) respectively. These publications have been cited over
12,000 times and over 4500 times (Google Scholar, accessed August 2015).
The NPK approach was originally presented in a single (multi-authored) book. The 3H approach
is presented in several edited volumes and special issues, and not all contributions easily ﬁt the
same mould (some are by authors who later vehemently criticised the approach). Etzkowitz and
Leydesdorff (2000), the introduction to one special issue devoted to the program, is taken as a
Perovskite Philosophy: A Branch-Formation Model of Application-Oriented Science
Critics of the NPK approach have pointed out that its sharp dichotomy between
two modes of knowledge production oversimpliﬁes both the history of science and
its contemporary diversity: some ﬁelds, such as pharmacology, might always have
revolved around Mode 2 knowledge; conversely, it seems an overstatement that
Mode 2 knowledge production is emerging in ﬁelds like particle physics. Relatedly,
the various aspects of Modes 1 and 2 are only loosely interconnected, which undermines the explanatory potential of the approach. Since Mode 2 must, for instance,
be both transdisciplinary and heterogeneous, possible cause-effect relations between
these phenomena cannot be coherently formulated. Finally, it has been pointed out
that the approach is biased: the focus on two ideal types, where one is explicitly
associated with a ‘traditional’ way of doing science and the other is diametrically
opposed to it, implicitly supports the transition to Mode 2 knowledge production –
in all the aspects identiﬁed (Godin 1998).
The 3H approach can be partly understood as a response to the NPK approach. It
does not posit a dichotomy between different ways of producing knowledge, with
sets of diametrically opposed characteristics. Instead, it emphasises that, in any process of research or innovation, there is an interplay between industry, government
and universities – the three strands of the helix. This interplay is in itself not a historical phenomenon, but any particular form that it takes is. Likewise, there may be
different “resolutions of the relations among the institutional spheres of university,
industry, and government” (Etzkowitz and Leydesdorff 2000, p. 110) in different
countries, sectors of industry or academic disciplines. Strategies that allow mathematicians to cooperate fruitfully with Internet service providers in China might not
apply to environmental scientists advising electricity companies in Germany. The
ﬂexibility of this approach concerns the relata as well as their relations. The historical forms of academia, industry and government partly co-determine each other. In
a “central-planning” form of the helix, for instance, the state controls both industrial
and knowledge production, and coordinates any relations between these activities;
this interaction itself in part deﬁnes all three strands. Furthermore, the strands and
their interrelations do not go through a random sequence of forms, but result from a
“reﬂexive subdynamics of intentions, strategies, and projects” (ibid., p. 112),
through which individuals and groups attempt to solve local problems, such as
attainment of research funding or improvement of competitive power over commercial rivals.
In addition to this general framework, the 3H approach offers an account of the
current “conﬁguration” of the helix. In it, the three strands overlap; they partly take
over each other’s characteristic tasks or missions, and hybrid organisations emerge.
Thus, we ﬁnd government laboratories and spinoff companies on high-tech campuses, and universities are encouraged to take up a ‘third mission’ of economic
development, besides teaching and research. The consequences that are outlined for
scientiﬁc research closely resemble the aspects of Mode 2 knowledge production.
Research is oriented towards particular contexts of applications, and needs to cross
traditional disciplinary boundaries. Unlike in the nineteenth century, new areas of
academic interest are not specializations within existing disciplines, but transdisciplinary ﬁelds like nanotechnology, artiﬁcial intelligence (Ahrweiler 1996) and,
more contemporarily, data science. In many of these ﬁelds, research projects involve
close collaboration with industrial partners, and occur in new “interstitial
communities” and “interface locations” such as digital media laboratories and science parks, rather than at universities and research labs.
Criticisms of the 3H approach have no canonical form yet, which may be a result
of its age and ongoing dynamics, as well as of ambiguities and unclarity in its central conceptions. Some critics (e.g., Shinn 2002) point out problems with the quasievolutionary terminology that is used in some of the seminal presentations. However,
it is at best unclear whether this terminology is needed to apply the 3H approach to
speciﬁc cases; many of the contributions to Triple-Helix volumes and special issues
avoid such terminology. Secondly, it has been argued (e.g., Mirowski and Sent
2008) that the 3H approach is driven by the idea that universities should take up the
third mission of creating economic value and should therefore become “entrepreneurial” (e.g., Etzkowitz 2002). In response, advocates of the approach emphasise
that the Third-Mission idea is conceptually independent from the Triple-Helix idea.
The latter indeed seems so general in its conception of industry-university-state
relations that it can hardly entail the former. A third type of criticism is that, where
the NPK approach may overemphasise differences, the 3H approach is overly
focused on continuities – both in emphasising that scientiﬁc research has always
been connected to industrial and state interests, and in downplaying any ongoing
changes in its overall story of perpetual change. This feeds suspicions that its advocates, like those of the NPK approach, tacitly approve current changes in academia,
but rather than appeal to revolutionary sentiments, they present a narrative of
continuity that is geared towards alleviating any worries regarding these changes
(e.g., Mirowski and Sent 2008).
The latter criticism has been formulated as an implicit commodiﬁcation of
knowledge, which might presuppose a neoliberal conception of the value of scientiﬁc research. It has been voiced against both approaches considered here (see also
contributions to Radder 2010). Apart from shared preconceptions, it also reveals a
shared level of analysis: both approaches focus on the overall organization of
research, both internally (e.g., in transdisciplinary areas rather than in hierarchically
organized ﬁelds of specialization) and externally (e.g., in interstitial communities).
This organizational focus is especially clear in 3H work, which offers in-depth
descriptions of how research in ﬁelds such as biomedical technology or nanoscience
was (re-)organised in response to various external incentives, without discussing
any of the content of research in these ﬁelds and how this might be different from
more traditional research. Similarly, although the NPK approach is, at ﬁrst glance,
about knowledge, it does not feature studies of central research activities (such as
theorizing, modelling, simulating, experimenting) that may be distinguished within
the encompassing category of ‘production’ and that have been the speciﬁc focus of
large bodies of work in the philosophy of science. Nor do the NPK and 3H
approaches thematise how contexts of application may shape the content of